PowerScale OneFS 9.2.1.0 CloudPools Administration Guide

CloudPools overview

CloudPools extends the capabilities of OneFS by moving data to lower-cost cloud storage. CloudPools can seamlessly connect to various cloud storage systems.

CloudPools is a licensed module that is built on the SmartPools file pool policy framework. CloudPools lets you move older or seldom-used data to cloud storage and free up space on your cluster. See Supported cloud providers and storage types for the list of cloud providers.

SmartPools provides granular control of file storage on a cluster. SmartPools uses a policy framework to segregate files into logical groups called file pools, and to store those file pools in specific storage tiers.

CloudPools expands the SmartPools framework by treating a cloud repository as an additional storage tier.

CloudPools uses a similar workflow to SmartPools. To store files in the cloud, you must define at least one account with a cloud storage provider. In addition, you must configure OneFS for cloud storage, and create file pool policies that identify and move files to the cloud. You can configure CloudPools to move files to the cloud automatically, based on file pool policies. Alternatively, you can use OneFS commands to archive individual files to, or recall files from, the cloud.

A SmartPools job runs the file pool policies and sends matched files to the cloud storage target. The job typically runs once a day. For each file that is stored on the cloud, a small SmartLink file replaces the original file on the cluster. To access file data stored in the cloud, users open the related SmartLink file through any supported protocol (SMB, NFS, Swift, or HDFS). Accessing data through a SmartLink file is called inline access. To fully recall a file from the cloud, use the isi cloud recall command.

File pool policies define the data to archive

CloudPools uses the same mechanism as SmartPools uses to define data to archive. You define files to store in the cloud by creating file pool policies. These policies use file matching criteria to determine which files to move to the cloud.

File pool policies are applied when the SmartPools system job runs. This job runs daily by default.

When files match a file pool policy that contains CloudPools actions, OneFS moves the file data of matched files to the cloud. Only metadata and a SmartLink file remain on the cluster, freeing up storage space.

Files remain visible in OneFS

Although file data is moved to remote storage, the files remain visible in the OneFS file system. CloudPools accomplishes this access by retaining a local SmartLink file. The SmartLink file is a pointer to the location of data in the cloud. You can read, write, archive, and recall files from the cloud as needed.

When a user views the OneFS file system through a supported protocol (SMB, NFS, Swift, or HDFS), SmartLink files appear to be the original files. When the user opens a SmartLink file, OneFS automatically retrieves and caches data as needed from the cloud. This operation is called inline access. Any modifications the user makes to a file during inline access are applied to the data stored in the cloud.

CloudPools provides a CLI command that fully recalls files from the cloud. Recall replaces the SmartLink files with the recalled files.